Classification and Extraction of Movement Rules for SMA-Net Robot

Abstract

This paper describes the method to evaluate the behavior of the flexible robot, called SMA-Net. In the case that the robot has the flexible structure, it is difficult to control and evaluated the behavior because of the high degree of freedom. Therefore, The first of all, we classified control signals for the purpose of investigating the relation between the control signal and the movement of SMA-Net robot using Self-Organizing Map. At next step, for the parameter sets from results of classification, we verified whether it is able to recycle or not as the movement rules.